Prostate cancer has been known to run in families for about 40 years and epidemiological studies have demonstrated an increased risk to close relatives of cases. This risk rises markedly when the closeness and number of cases in a cluster increases. There has been considerable debate about the genetic model, in particular whether there is a commoner lower penetrance (moderately increased risk of the disease due to the gene(s)) in addition to contribution from high risk genes. For the first time, molecular results are starting to emerge, indicating the location of high risk genes. These have shown that there is evidence for more than one site of a high risk, gene two sites on chromosome 1 and one on chromosome X. These do not account for all clusters of prostate cancer cases and further genes remain to be discovered. This article also outlines the contribution of the numerous collaborators in the British Prostate Group to the UK Familial Prostate Cancer Study.